London South East 00:01
You are listening to investing matters brought to you in association with London South East. This is the show that provides informative educational and entertaining content from the world of investing. We do not give advice so please do your own research.
Peter Higgins 00:17
Hello and welcome to this Investing Matters Podcast, today I have the huge privilege of speaking with Alistair Unwin, the deputy portfolio manager of the UK FTSE 250 listed Polar Capital Technology Trust (PCT) with a market cap of 2.95 billion as we record today.
And the trust is up 40.6% year to date and it's also running a 10 years CAGR of 17.83%. Alistair, welcome to this podcast with me.
I just to say before we start to qualify your expertise, you've got a first class Honours in History from the University of Cambridge, you're CFA chartered holder, and you've got 12 years experience in the investment industry. So thank you ever so much for sharing your insights with us today.
Alistair Unwin 01:14
Not at all, my absolute pleasure to be here, Peter.
Peter Higgins 01:18
Thank you very much.
Now, in this Investing Matters interview Alistair, I want to start with just laying the groundwork for our investors that you are one, you know, amongst your team as well, of the experts in the artificial side of things regarding Polar Capital Holdings, artificial intelligence, essentially been driving the outperformance of the Polar Capital Technology Trust year to date Alistair, I want to go back a little bit and just lay some groundwork here.
And I want you to share with me if you would be so kind to start sharing your journey from Deloitte in 2005, as an audit associate, to being a fund manager at Neptune Investment Management concluded up to about November 2018, just give us a synopsis of that journey, all of your learnings as please?
Alistair Unwin 02:08
Course and yes, it's been a wonderful time, we've been very, very fortunate to be tech investors during this period, because it has been as I'm sure you and many of our listeners be aware, it has been boom years for tech for all kinds of reasons we can go into.
But in terms of my own career, I started in audit for Deloitte, which I enjoyed but found it a little bit dry.
My father is an auditor, so I'm I can sort of be rude about them.
And then I moved to a company called YouGov who do political polling straight after university, which I enjoyed tremendously, and started to do more work on the data side and we did some work with Bloomberg.
And we did some more financial markets facing stuff, which I enjoyed very much.
And then I made an unusual move to one of YouGov’s investors from being a just an employee of the company to one of their shareholders, which is Herald Investment Management, which is run by a lady called Katie Potts, who you may be aware of is a legend of the tech investing game and the UK investing game.
And it was my great privilege to work there for five years or so. I then moved to Neptune to launch a tech fund, which did quite well.
And I took over a U.S. all sector fund and kind of versus the S&P fund while I was there, but that tended to have a tech weightings that did quite well, given how well tech was doing. I've then a brief foray at a smaller financial startup before moving to Polar in whenever that was sort of June 2019.
Peter Higgins 03:43
Yeah. So we're up to June 2019 and you gain the job.
Could you just tell us how that came about?
Obviously, they’d been tracking it regarding all the things that you've done prior to that.
So June 2019, you’re a fund manager, you arrive at Polar Capital Holdings.
How did that come about and what we offer sort of role within the company there?
Alistair Unwin 04:06
So just to set the scene, Polar Cap Tech is the largest or one of the largest tech teams in Europe, it is one of the best place it will most sort of we can say that about ourselves kind of highly regarded long standing basis doing tech in Europe very successfully for some period of time.
And I knew Nick Evans and Ben Rogoff, who managed the franchise at that point, they managed to trust and Nick manages the users’ product.
And I done trips with them and sort of got to know them over the years. I've always been keen to work here, frankly.
And eventually an opportunity came up as they're expanding their team. And I came on initially to cover software and fintech because the tech space is sort of ever growing in its reach and capacity.
And they felt they needed more more heads to really do justice. I guess we've had another couple of analysts over the summer actually, we think that that's getting even more of a case with AI.
That's what I started here. And then obviously went straight into COVID, which was something of a challenge, particularly on the team you've just joined.
But to be honest, given that we spent a lot of time traveling and seeing companies I mean, we were set up on Zoom and RingCentral and things before COVID, so for us it was obviously a big change in lots of things, certainly changing the investment environment, but practically it didn't make all that much difference given we were pretty well set up for remote work already.
Peter Higgins 05:30
Brilliant, thank you.
You touched on how large the Polar Capital and management team is and the scale of the assets under management there.
And the team won best fund management group for the category of 20 billion euros to 100 billion euros category, the Funds Europe Awards this year. So still going strong.
Alistair Unwin 05:53
Yeah. I mean Polar Capital was founded on tech on the tech franchise by Brian Ashford-Russell and Tim Woolle, who came out of Henderson with the trust as it is, or in its current form is PCT.
And so there's a it's really at the core of Polar.
We're about 40% of the assets, I think. And there's some other fantastic funded franchises, as well.
But tech has been very much core to what we do and yeah, it's just at the heart of the business.
Really, we think having a big team is important, because it's an enormous space technology.
And it changes very, very, very quickly. It's not like looking at some other sectors where it's the same old faces, you know, there are new companies coming to market.
I think in 2021, certainly, we had 10s of investable companies coming to market and you need a big team to really do them justice.
In our view. The challenge to a big team is that when small cap tech is not doing as well as mega cap tech, or, or very large cap tech, you've got a portfolio in PCTs case of sort of 90 to 110 stocks.
And the marginal stock if it's a small cap is less likely to do well than mega caps. And that that's been a sort of headwind to having a larger team.
But we're sort of hopeful we might be past the worst in that we can really kind of show what a bigger well, resource team can do.
Peter Higgins 07:15
Indeed, you've almost covered my question, but I'll get you to go over it a little bit more here reiterate some more points to give us an overview of the Polar Capital Technology Trust team and also its investment philosophy, please Alistair?
Alistair Unwin 07:30
Yes, so the team is essentially we have three products in the team and the trust is the second largest of them.
So we use, it's probably it's about five and a bit billion dollars.
The trust as you say has about four and then we have an artificial intelligence fund as well, which is about 400 or 420 million or something like that in dollar terms.
And each of those products has a lead manager so Nick Evans leads on the user product Ben leads on the trust which I'm the Deputy Manager on and Xuesong Zhao leads on the artificial intelligence fund.
And then beneath that are Fatima Yiu and myself who both sort of assist with the management of those, Fatima more UCITs to me more on the trust.
And then we have five analysts five full time dedicated analysts beneath that largely split by sector of sub sector I suppose within the tech space.
So for example, I cover FinTech and application software. Fatima covers infrastructure software, cybersecurity software, Paul Johnson covers video gamers and EVs for us and so on and so on.
And we believe that you need domain expertise in technology, because you have to get quite detailed because the debates around stocks are quite detailed, and what drives stocks is sector specific.
Obviously, rates have a big impact on some of the growthy areas, the market and the whole market valuations as a whole.
But you really do need to know the detail in tech that they're not equivalent assets, like choosing between different regional banks, or different insurers, where it's on a kind of well-known set of metrics, basically, tech, everything is always up for debate, tech companies can be very, very good articulating their story. So you need a kind of healthy degree of skepticism with that.
But there are also moments where they don't know and you don't know, as an investor, quite what's going to happen.
So there are moments where you need something that is much more like a belief that this company has differentiated technology, even if the business model maybe isn't as fully fleshed out as you might want to, or there is a potential competitor who you think is going to become a big issue. But whether it's going to be fatal is the sort of thing we spent a lot of time focused on.
Peter Higgins 09:46
Indeed, and that's prove this point this year, because after what was a difficult 2022, during which many tech stocks were hit hard, many market commentators and analysts were very pessimistic about the prospects for tech stocks going into 2023.
Firstly, please, would you explain in layman's terms, artificial intelligence, followed by generative AI, then inform our global audience what happens in 2023, to transform the fortunes of those very tech companies, which analysts were so pessimistic about?
Alistair Unwin 10:20
Well, if I could take the second one first, if that's alright, because Gen AI is the sort of the AI story is that kind of finishing better bit at ‘22 was a very, very challenging year for risk assets generally, but particularly for growth tech equity managers, or growth equity managers like ourselves.
And the primary reason for that was the Fed raised interest rates significantly, we saw a significant compression in the valuation multiple that investors were willing to pay for long duration assets.
So that's assets where a lot of the value is in, out way out into the future, very sensitive to the interest rate used to discount those future cash flows.
So that was a big headwind, and we saw something like if you looked at take a Tesla or something like that as a kind of proxy for that valuations cut by about two thirds, that was really, really tough.
On top of that, coming into ‘23, everyone knew we were going to have a recession. It was just a question of when and how bad, tech had had a great time during COVID.
I mean, everyone at home buying things online, every company needed to get on Zoom and improve their tech estates, sort of almost ideal environment for tech.
How can it be any better? You've had the sort of top of the tech run and other sectors are going to do better and text not going to be bad.
It's just not going to be that interesting. And so that's how we came into the space and then this year, what's happened, firstly, the macroeconomic environment has been better than people thought growth has held up better driven by a very strong labor market, resilient consumer, particularly US consumer, but also in Europe as well. And there hasn't been a recession, there's global recession that is supposed to exist just hasn't turned up.
So that's been really the first thing. And given how negatively people are viewing growth equities at that time, that definitely made a big difference.
The second bit was, you'll probably remember in 2022, we were tracking layoffs in the tech sector almost week by week, and there were headlines about Meta’s cut 17,000 people and Google's cut 10,000 people.
And as you tracked it, but you could see how literally how many people were being laid off week by week or day by day.
What that meant, of course, is if the macro-economic environment doesn't deteriorate, all of those companies who've laid people off see the margin benefit over the next year, which is what we've seen in 2023.
And you've seen that at the very big companies. But you've also seen it further down market cap spectrum.
So if you take Software as a service stocks, which are the sort of growth your area were criticized for never making money, they took headcount reductions, they reined in advertising expenses, they did the things you'd expect to see coming into a downturn that never arrived, therefore, that you've seen their margins increased significantly this year.
So we've had kind of better macro, better profits or better profitability for a sector that has been criticized for lack of profitability, we've always been a little bit more sanguine that tech can basically print the margin it wants, it's within its gift to do it, you just have to give up some growth.
And then the final bit, which is the most significant, I suppose, from our sector, specifically, is the emergence last year, November was probably two weeks ago, last year of generative AI, and this extraordinary new technology, which we are still in the very infancy of which we believe is we and the market can I mean the performance of the stocks as yours as the market is of this view, to some degree as well, is going to be transformative, both for the sector, but actually for other sectors outside tech for humanity as a whole.
We don't think it's overstating it to talk in those terms.
And what this technology is at a high level, is that artificial intelligence is teaching a computer to think and learn like a human.
That's really and what that means is that anything that we would normally associate with human intelligence, reasoning, problem solving, understanding natural language, and producing natural language, particularly those things, if they can be simulated to a human like or superhuman level, using a computer that is artificial version of human intelligence, therefore, artificial intelligence.
And so that's how we kind of think of it at a high level. And that has been transformative, because we can suddenly address human knowledge work with tech in a way that we haven't been able to before.
We think of the tools we use as ways of getting our thoughts down on paper or on the screen or a spreadsheet to help us do some faster or produce models or whatever it might be.
We're now getting to the next generation of this were really generative AI in particular, reflects our ability to ask computers to do things for us. And I mean, computers very broadly, not just PCs, but mobile phones, and even computers that have ambient computing.
And when you start to think about that, you look back at prior examples of big technological inflections.
For example, when the PC came about the new user interface was the GUI, the graphical user interface.
And once we had the GUI with Windows three, in 1990, the number of PCs explodes, a number of use cases you can do with a PC explodes, we go software market, and we start to see the networked enterprise and all the things you'll be familiar with from the, through the 90s, that great ICT revolution, as it was called, then, that was down to this or in part to this new user interface.
Same thing happened with the iPhone, there were 300 million smartphones before the iPhone in the world.
But what you could do with them was quite limited. It wasn't that they didn't work, it was just the user interface, you had to scroll through lots of little menus, it was a small, not very easy to see screen, the iPhone turns up with a touch interface, a new way of just sort of prodding the computer to do what you want to.
So after that, we then see the emergence of the App Store.
We then see the emergence of businesses like Uber, like Airbnb, like Facebook, and all of that extraordinaire innovation on the top of that new user interface.
Fast forward to what is generative AI, we have now got a user interface that is quite simply our own language.
So we can talk to a computer, whether that's by chatting in the chat box type Chat GPT, or via speaking to a smart speaker, it doesn't really matter.
We can now tell the computer in our language, not in code. What we want the computer to do for us do this analysis for me send this email, summarize this conversation, find me somewhere to go on holiday, create me an itinerary.
The use cases are literally just beginning. But that inflection point we think is very, very much here and we've kind of pivoted the portfolio to take advantage of that.
Peter Higgins 17:06
Fantastic.
Thank you for that for reply. I appreciate that Alistair.
What I'd like to do now at this stage, please is to drill down on this specific It AI stocks that you and Polar Capital Technologies trust team have within the portfolio, its applications and characteristics that led to them being selected.
If that's possible, I'd like to start firstly, with the sector exposure you have to software, which makes up 27.1% at this particular stage regarding sector exposure, can you give us the stocks you've chosen for that particular area and why?
Alistair Unwin 17:43
Well, we have to be quite careful talking about specific stocks as a fund manager when we're active in the stocks quite frequently given for retail audience.
But I think at a high level, you're absolutely right.
That is our software exposure, as it sort of appears on our fact sheet.
The way we're actually thinking about the portfolio increasingly, is not so much in terms of the sub sectors, but actually in terms of how each holding relates to AI.
And where does it fit into that and so the way we've went through a process earlier this year, that's a sort of ongoing process, looking at, well, how does the portfolio rate AI?
How will we, what is the best way to invest in this? Is this opportunity amazing?
Next stage, how do you invest in it, and we think that we've got about 60% of the portfolio is in AI enablers. And that split into roughly half into the cloud computing companies so at the very high level, that's the Amazon's AWS and Microsoft Azure and Google GCP.
And then the other half is of that 60% is semiconductors.
So there is a new computing stack, as its referred to, which is basically their component tree, the chip, the design tools, the memory, all of the different aspects that go into that that sits below out that is different from the computing stack that was used.
And it's not massively different. But the implications for individual stocks, particularly smaller stocks, particularly in markets, where maybe we haven't had as much exposure in recent years, like Japan or Taiwan.
When you move to a new stack, it's all up for grabs again.
And so we think that there is some interesting opportunities in some of the Japanese small cap areas, for example, in companies that we might not have, that might have been too low growth on not that interesting for us, if they are a play on or have exposure to changes that happen because of generative AI, and an AI computing stack, then they become more interesting.
And that's where we're spending a lot of our time hunting. So that's 60% of the portfolio.
Another 20 is we believe in companies that are beneficiaries of AI. And that's to say they are able to generate incremental revenues or improve their competitive position using generative AI that they weren't before.
So we're now in sort of 80% of the portfolio, there's another 15. And really most of the remainder, which is adopters of AI.
And that would be companies in areas like FinTech, where we have we don't have very much but they will be using AI often have been using machine learning like pre generative AI for some time.
And those companies we don't count as part of our sort of our exposure, but some people do and some people actually how much we've got anyway. So that tends to be how we think about it.
Peter Higgins 20:38
Thank you very much. I will point out at this stage that we do have institutional listeners on that come to listen to the Investing Matters Podcast as well.
So that's why I'm posing some of these questions, within your top 15 holdings, Alastair, there are some holdings, which our listeners I don't think will be aware of and one of those is, I would say a mega cap really, Arista Networks, 69.7 billion market cap company whose shares are up 84% year-to-date, which is just staggering.
Could you tell us what were the characteristics behind that particular stock?
And what led to that inclusion regarding that one, because it's still remarkable for me that institutions in the UK are not aware of companies that that sort of size, nevermind the retail investors.
Alastair Unwin 21:34
It's a funny one, one of the things that's happened over my career is what counts as a small cap has changed quite significantly, you know, a billion dollars, or a billion used to be small.
That's micro now in tech, you know, you need below 10 really is a sort of small cap in profit terms.
And because the sector has eaten the profits and other sectors, I think is probably the what's happened there.
And it's no surprise, but Arista is Arista is a good example.
So a risk to make switches primarily or an operating system for switches and switches.
So they are a networking company, probably the leading cloud independent cloud networking equipment provider, and their big customers are Microsoft and Meta or Facebook.
And they specialize in Ethernet, which is one of the networking standards.
And really, it was founded by a group of people out of coming out of Cisco, who very, very smart people, many of whom had already had successful careers in in prior companies, and even before Cisco in some cases, and they wanted to really just rebuild things for a cloud world.
And if we do this again, how would we do it? They really took that kind of approach to building a networking operating system, and then selling switches on the back of that.
And then they signed up some of these very big cloud customers, and increasingly, they're going into the enterprise Isn't the campus and you know, trying to win the other bits of Cisco's business that they haven't won yet.
And they've been very successful. And we've been shareholders to varying degree for many years now, the opportunity for them on the AI side is while the first generation of AI so we think of that the training that's happening at the moment, we're training these large language models like GPT Four that sits beneath Chat GPT, or Gemini, which has just come out of Google, those will be trained using Nvidia chips.
And they will be trained using generally networking from a company that was called Mellanox, which is a standard that which is now part of Nvidia.
So Nvidia acquired that company a few years ago, so in Nvidia selling chips, and they're selling a lot of the networking at this time, when we what yeah, what a business.
And then when we get to slightly, probably a couple of years’ time, the expectation is that generally, open standards tend to win over closed Ethernet is an open standard Arista and a number of other Ethernet players, including Cisco and to a lesser degree, Juniper will have a play in AI as you need to connect all the different clusters together.
And that will then be essentially, the more we're doing with data, the more data you need to move around, the more network capacity you will need and that is ultimately demand for Arista and other networking companies’ products.
Peter Higgins 24:24
Thank you for that.
Now another holding you have CrowdStrike Holdings, $58 billion, behemoth shares, actually moved 130% year to date.
Now we all hear about these cybersecurity companies, we know we need them, there's lots of stuff going on, and put some, you know, databases, infrastructures at risk.
What's so different about CrowdStrike and what they're doing regarding AI, that makes them stand out amongst all the most of their peers should I say?
Alastair Unwin 24:55
Well, they along with another company called Palo Alto Networks, which is an even bigger company, I think that the largest in the world, they are the largest cybersecurity independent company.
I mean, Microsoft also has a huge cybersecurity business, but you can't see it broken out because it's within even bigger business.
It they've been beneficiaries from the consolidation of spend, actually.
So typically, the way it has worked is that you get a new cyber threat, a new company, or in fact, a series of new companies are funded by venture capitalists to go and solve this threat.
They do very well with that there's a buying cycle. And then they get acquired by one of the bigger companies, often by Cisco actually, or similar companies to that.
And then that's sort of the end of it and the VCs take the money, it's been a great business.
Increasingly, as you move to an AI world scale appears to be mattering more, because you need to be able to see all the threats in your network.
And then when you see a threat in one place, or in one jurisdiction or one type of threat, you can then roll that out across your estate basically and protect everyone from it.
So CrowdStrike do next generation endpoint, again, after the sort of the, the old 90s winners in that space. And they have been taking share along with Palo Alto for many, many years. Now, the AI angle on cyber is going to be very, very interesting, because one of the unfortunate consequences of this incredible new technology is that the bad guys will adopt it too.
And the threats will get dramatically more sophisticated or already are getting dramatically more sophisticated.
Of now the good the good news, if it is good news, it is if you're a salesperson at CrowdStrike, is that the tools to defend you getting more sophisticated at the same time.
Cybersecurity is a sort of an evergreen theme in the portfolio if you'd like because the threats get bigger and bigger as more of companies own.
And then tech companies’ own value is tied up in their data or their IP or their technology.
So the need to protect it goes up as well.
And as we moved I mean CrowdStrike is big success was as we move to a cloud first world, they were better positioned than the on premise incumbents to defend endpoints.
And that's really been the kind of big need they've met over the past five or seven years. But we think cyber remains at a pretty attractive market.
And other stocks have obviously, as you mentioned, had a good run. And they often traded quite expensively particularly relative to the growth but it is an area still of secular growth where the larger players are consolidating share, and you can see with AI coming that the potential for that to continue it looks reasonable.
Peter Higgins 27:42
Thank you.
I love the fact you've touched on a very important point there and I'll just paraphrase what you said.
That is part of it and I’ll reframe this question for the winner take most revenue and profits trend continue for the ‘magnificent seven’ and the other AI enablers, do you think longer than the market naysayers and shorters can remain solvent?
Alastair Unwin 28:08
That's a good question.
We've seen this across and it actually happens outside of tech as well.
But we've seen this phenomenon of the strongest getting stronger and returns to scale.
Well, in tech and elsewhere, just remaining persistent.
I mean until very, very recently, small cap tech is it sort of 20-year lows versus large cap because being subscale being a point solution is getting harder and harder, not easier.
And you see it in funny areas, like the proportion of money made by the very largest music artists is going up as a percentage, even as you might think the internet would democratize it.
It's been a fascinating phenomenon and it's to do with network effects.
And it's to do with these increasing returns to scale and the productivity growth that can be driven by those superstar companies and the kind of superstar economics, I guess, power law distribution of returns.
We haven't seen much evidence of reversion yet.
Actually, we are, AI is entering one here, because one of the reason those magnificent seven companies have performed well this year, is that, yes, they were becoming more profitable.
Yes, the macro held that better.
But they are also to a varying degree, pretty well positioned for AI I mean, if you think to be a winner in AI, at least at the early phase, you need lots of data, lots of computes lots of very smart technical people, and a means of distributing it, then the magnificent seven all look fairly well positioned, actually.
And then we can debate valuation.
But really, then it's hard to get AI is necessarily going to be bad for them.
Even if you want us to take a skeptical view, given how well positioned they look to me, you know, they have GPUs, they have data, they'll have proprietary data at a scale that other people don't.
If AI is in the business of extracting insights from enormous bodies of data, then you take big companies, you have the biggest data, and the ability to extract those insights should be well positioned.
The pushback to that, I think, is that when you get one of these tectonic shifts in technology, as we've happened with the internet has happened with the mobile phone, probably to a lesser degree in the cloud.
Typically, the incumbent doesn't win most familiar with Clayton Christensen's book, The Innovators Dilemma, that is a classic and it's a brilliant book, I would recommend it to everyone understanding why the incentives for an incumbent to disrupt themselves, if that's the right term, are not there.
But it's a bit unknown, pointing out exactly how generative AI is going to undo Apple's dominance of the smartphone industry is a speculative exercise at the moment, it doesn't mean it can't happen.
But it is a it is a speculative exercise, what I would say is that the time at which Apple's dominance of the smartphone industry, for example, could be undone would be at one of these inflection points, where other people do have a window, whether it's possible to make an execute on it, they can get funded and all the rest of it, but you are stuck, there was a company called Humain, who have launched a pin, which is a sort of voice activated product that will use a be a sort of individual personal assistant, that will potentially be an iPhone Killer.
I mean, but we're talking highly speculative, but at least there are maybe the beginnings of more credible disruptors to sell to a company like Apple, who has just proved phenomenally successful at defending that core position or core market position and core share of profits of the industry actually.
Peter Higgins 31:56
Brilliant.I love that reply. Thank you ever so much for that insight on that other company as well, Alastair to now you touched on the valuation side of it.
Now I need to ask you this regarding PCT, what are your selling criterias and discipline for the PCT portfolio in the sense of valuations or see the incumbents lunch being eaten by a new company coming into the space?
Alastair Unwin 32:24
So the part of the market, we focus on as we deal with private.
So nothing... we leave out to VC it's very low hit rate very high return.
That's not what we do, we don't think that the trust structures is the right one for that.
We also steer very clear of incumbents, which would be I've been Cisco does seem to be picking on Cisco today.
But that would be a company that would not held for many years.
It's not to say it's a bad company, necessarily. It's you know, it's got financial metrics can look fantastic.
And many of these companies indeed do look fantastic on financial metrics.
We avoid them because it's our contention, I think it's the evidence backs it up that change in the technology, space is nonlinear.
So when you go from being a winner, as Cisco undoubtedly was, for many years, to an incumbent, the business doesn't fall apart, but you become much less interesting as a technology investment,
because you are forced to do M&A that carries more risk, you are forced to lever up and buy back stock, you're forced to become a sort of defending your turf rather than going into new markets.
And it's really hard to have a second act is the other thing.
And very, very few companies have done it. Microsoft is the most notable exception, and then having done really well in the cloud this year and now potentially even leading AI.
So it's not impossible, but that's a big part of how we how we think about things.
So what that means for our sell discipline is if we think that a company is moving from being a disrupter to an incumbent, our sensitivity or our need to hold that company goes down.
And that may not be at the point that it's just massively expensive, because companies can appear expensive, but then grow for a sustained period of time at a higher rate.
Which means that when you look back in three or four years’ time, they weren't actually that expensive.
I mean Nvidia is a good example this year, where you've just seen enormous earnings upgrades and just incredible growth. The beginning of the year, people pointed at the P/E and said, it's expensive.
And you know, if you didn't do it, we're going to move you were right.
But it's kept us out of trouble actually avoiding hanging on to companies that are ostensibly fine.
But we think past the best of their growth years, and their incumbents, and then we're probably beating a retreat is the best way to think about it.
So that's how I would frame ourselves discipline, as opposed to, once it gets to this time sales, we always trim it or anything like that, because it's very, I think the if there's one common theme for our approach is the avoidance of hubris.
Because this this sector is, or it can lead investors to be very hubristic either in terms of how much they think they can know, they can understand or predict the future, or how much the degree to which people can know the right price of an asset.
Because you can see, I mean, even the last three or four years, you've seen what people will pay for a unit of growth, or a unit of profitability, swing around wildly with the moves and rates and the moves and things that have nothing to do with tech.
So the hubristic, we think it can be hubristic to say, well, I know that at 10 times earnings, I buy tech stocks, and I always sell them at 30, or whatever. That's not really how we think about it at all, actually.
Peter Higgins 35:49
Brilliant.
Thank you. You've touched on hubris there, so I’m going to ask this question at this point, one always asked to look to avoid overconfidence.
Therefore, what do you personally see as the greatest risks and challenges to the proliferation of AI?
Alastair Unwin 36:06
That’s a very, very good question.
I think the challenge of regulation, the necessity of regulation is here.
And I think there's no world in which this technology is not is not regulated, how companies and we have how we saw the EU last week gleefully announced that they put out the AI act to regulate AI already, which even by their standards is very, very quick.
And then we think probably the other country will very likely other countries follow how companies deal with that regulation and the approach they take is a risk.
Could that slow things down?
Absolutely, you have some unusual risks around governance, as you saw with the OpenAI debacle, where someone was sacked, and then had another job and then was brought back in it was based on five days, which is pretty good going.
And now Microsoft is being investigated for whether they have control, which I think it's worth, it doesn't look like they don’t have risk control. But anyway,it's real, the regulatory risk is real. So that's an area.
Another area is that the big breakthrough for AI, or for generative AI was the foundation model, which is what came out of Google in 2017, in a white paper.
And the key stats I think to know is that pre that paper coming out, if you were using non transformer models, the amount of compute you could use to train them increased about eight times every two years. So quite a lot.
After that paper coming out and using techniques like self-attention, the ability to process in parallel, and some other innovations, you increase the amount of compute, you could use to train by 275 times every two years.
So just a massive inflection in the scale of models you could produce, which is why you can train a GPT, three and a half on the whole internet, rather than just a sub sector of it, where you get to this very, very different scale around these emergent properties, they are able to generate an understand natural language that can generate images, we just move to the next phase, if you like, the risk from a technological perspective, and there's not much evidence this is happening, by the way, would be that that rate of innovation just slows, we just reach some natural frontier where actually the next incremental innovation or the next wave of compute that you're able to use doesn't make as big a difference as the last one, there has to be a possibility, we can't really know that.
The other major area of risk to the AI story is around Taiwan.
And given that 93 or 94% of leading edge chips are produced in Taiwan, if there was some escalation of hostilities or some issue with production there.
I mean, they have water shortages over time, for example, that would be probably not fatal failure story, but would certainly set it back pretty substantially.
Peter Higgins 39:02
Thank you for that for reply. I appreciate that.
Now, I want to just break a little while here briefly, and ask you to share with us on our global audience, Alastair.
What's your personal investing strategy? How do you go about investing your own wealth long-term, your savings and investments? Are you all in regarding Polar Capital's bonds and trusts? Or do you have some diversification elsewhere?
Alastair Unwin 34:37
Diversification within Polar, but I don't have very much outside of Polar.
So we have some of our pay is held back as deferred compensation.
And then we're required to invest that in paper stock or pay the funds.
So necessarily, I have a lot of that independent funds. I also have some kind of funds outside of that in my sort of ISA and my son's JISA and that kind of stuff, because I work with the people here and I think they’re phenomenally capable and very hard working.
And so to me, that makes sense.
We've got enough diversification across different products.
So we have a healthcare team and insurance team and financial team and an Emerging Markets team.
So you can see that there's, there's enough diversification for me there. So that's the bulk of it.
Peter Higgins 40:09
Brilliant. Thank you.
Now, I want to touch on this.
I was quite pleased to see this when I was looking through some of the notes that I was looking for regarding yourself.
And given the cost of living crisis and food poverty.
I'm an ambassador for charity myself, so I'm thrilled to see Alastair that you and Polar Capital Holdings, supports a food bank, please share a little bit about that?
The one that you support and what support that Polar give you regarding that as well?
Alastair Unwin 40:36
Yeah, so Polar does a lot on the charitable side.
We also have a relationship with a local school, which has been very, very successful.
We've sponsored some of their students through university, some of them come and work at Polar.
It's been some movement, a lot of thought and effort into in the last couple of years.
And Polar has a really, really nice thing every year where members of staff nominates particular charities or things that that are sort of personal to them.
And then Polar often allocates as a committee at Polar decide where to allocate money, but they allocate generously to those charities, which is sort of the food bank on my sister's very, very involved with running it.
So it's a sort of family affair. And we host quiz nights, which my brother is the MC all the time.
So we do we do quite a lot with them.
But it's yeah, shame it’s required. But it's been something that we've been quite involved in for some time.
Peter Higgins 41:35
That's fantastic, keep up the good work, I love that. Thank you very much.
Now, I've got a final question for you, Alastair and this is one where I just want you to qualify the reasons why it's important to invest long-term essentially.
So launched in 1996, Polar Capital Trust has survived and thrive despite experience in the dotcom boom and bust, 2007/8, a great financial crisis.
And obviously, most recently, the COVID 19 pandemic, Alastair, what do you and the team see as being the analytic traits and strengths that has led to this enduring success of the PCT Trust, which is currently compounding 10 year CAGR at 17.83%?
Alistair Unwin 42:20
I think that's a testament to the people who worked here before me more than anything else, particularly Brian Ashford-Russell, and Ben Rogoff, who had been the team managers, since the trust, as it was in that form, and our bridge aims to, I guess, bring the best actually of what the benchmark has to offer.
So we run it in a benchmark aware way.
And that's partly avoiding hubris, you want to tether yourself to a good benchmark, not a bad benchmark, you want good competitors, or bad competitors, whichever keeps you sharp.
And I guess a belief in the non-linearity of tech change the idea that you can only know so much, and you work really hard to make sure you know, as much as you can, but there is a limit to what you probably can know particularly around how to value things, because that changes over time, the importance of diversification, you know, we're at 90 to 110 stocks in the trust, that leaves us well diversified things do go wrong.
And we have, you know, we have stocks that are down a lot every year, but that normally confined to the portfolio tail.
But if we're doing our job well, and then tethering yourself to a sector that has been able to innovate, and has actually delivered the growth and the profit, and it's one of the push backs against the dominance of protecting the benchmark of 28-29% of the S&P, it's about 20 of the profits.
Now we can debate whether that's cheap or expensive, but it's not, we're not in a situation like 2000, where we get to sort of two and a half times the market multiple and it's all a bit untethered from reality, and there's no profits, and there's no cash flows.
It's very different.
So I don't think it's all of that all of this together, and then absolutely loving the space. I mean, this is a wonderful space to invest in. It's with Mike and the team's great privilege to do that. And we hope to do for many years to come.
Peter Higgins 44:14
Fantastic reply.
Yeah, I can see clearly that you know, from being an audit associate that you're far more excited in the tech space Alastair -great.
Now, ladies and gentlemen, thank you ever so much.
Wherever you are in the world for listening to this Investing Matters Podcast, it's been a privilege speaking with the Deputy Portfolio Manager, Alastair Unwin, Fund manager at Polar Capital Technology Trust.
Alastair, thank you ever so much, love to the team and we'll look forward to speaking with you again and the team in 2024.
Wishing you continued success.
Alastair Unwin 44:53
Thank you very much. Thank you.
Peter Higgins 44:55
Take care, god bless.
London South East 44:58
Thank you for taking the time to listen to Investing Matters. Be sure to check out the London South East website for free tools and info to research your next investment. You can also join in the conversation on our social media channels. And don't forget to subscribe to our YouTube channel for more content, including our CEO interviews. Catch you next time.
